Introduction
This work is part of a wider research project focused on understanding the role of family relations in building a moral attitude in pre-adolescents (Malagoli, Ardone, 1993).

The systemic approach which is followed in this paper, clearly shows how the behavior of a pre-adolescent (e.g. his process of acquisition of moral rules) cannot be examined without taking into account the affective components of family life.

Goal
Analysing the various mechanisms leading to transgressive behaviour (according to Bandura's definition of it), we tried to find the connection between transgressive behaviour and family types, as they are subjectively perceived by the families of our sample and represented in Olson's Circumplex Model (1982,1986,1991).

More in detail we aimed at verifying the connection between the emotional relationship in the family and the transgressive behaviour in families with pre-adolescent children (11-14 years).

Hypothesis
Our hypothesis is that some family types could be considered, more than other types, as factors contributing to moral disengagement of pre-adolescents.

Sample
The sample consists of 360 families composed of mother, father and one pre-adolescent child, amounting to a total number of 1080 people.

Instruments
The Moral Disengagement Scale (MDS) is an instrument used by Bandura to measure the functioning of the internal mechanisms regulating moral behaviour.

In our research we used the Italian version (by G.V. Caprara) of this instrument after the necessary validation.

The scale is composed of 8 subscales related to 8 mechanisms which are typical, according to Bandura, of moral disengagement. The FACES III(Family Adaptability and Cohesion Evaluation Scale) was used in the Italian version and in the two versions Real Family - Ideal Family.

This instrument consists of two subscales (Cohesion and Adaptability), each of 10 items. Cohesion applies to the emotional bonding between family members, Adaptability concerns the ability of the family system to reorganize in response to situation and developmental stress.

Results
The data was subject to different levels of analysis.

At a first level we proceeded to correlate separately the scores obtained on the two dimensions of FACES to the scores on MDS by the three different groups (fathers, mothers and children).

The scores of cohesion and adaptability for the version real family, show a negative correlation with the scores of moral disengagement scale. Higher levels of cohesion and adaptability perceived in one's own family correspond to lower levels of moral disengagement. Similar results are obtained for the version ideal family.

Concerning the classification according to family system types, for the version real family the majority of parents belong to the flexibly connected type and the majority of children to the flexibly separated type.

For the version ideal family the fathers maintain the same type, while the mothers move towards the type chaotically connected and the children towards the type flexibly connected.

Distribution of fathers, mothers and children in the types of the Circumplex Model and scores on MDS

By an analysis of variance we correlated the scores obtained on moral disengagement to those obtained on FACES III for all the types of this model.

Concerning the children the results of the analysis of variance show a significant difference (p=.01) between the family types and the average scores of moral disengagement.

The children having the perception of a family structure belonging to the flexibly disengaged type and to the structurally disengaged type get the higher score on MDS.

Both these types are characterized by a very low level of cohesion; transgressive behaviour of preadolescents is strictly related to the perception of a low level of emotional support between members of the family.

Children of the chaotically connected type, characterized by a higher level of cohesion and adaptability, show, on the other hand, a lower tendency to transgressive behaviour. A high level of cohesion between members of the family tends to facilitate the control of the mechanisms causing the transgressive behaviour.

Conclusion
The hypothesis of a negative relation between the scores on Cohesion and Adaptability and the scores on Moral Disengagement seems to be confirmed.

There is less tendency to moral disengagement when the family structure is perceived as more connected and more flexible in rules and roles.

A strong cohesion between members of the family is a psychological and relational dimension strengthening the self-regulatory mechanisms of the children, hence reducing their tendencies to transgressive behaviour.

A high level of family cohesion and flexibility of family rules help the parents to have a more positive view of the pre-adolescent child.

In conclusion, we think that the pre-adolescents belonging to a disengaged family - due to the poorness of emotional bonding, the scant reciprocity in the relationship between members, the flimsiness of parental functions - miss the necessary support for the acquisition of moral rules.

R. G. Ardone, S. Vicini: "Pre-adolescence: family structure and moral disengagement". Paper presented at the First International Congress of Adolescentology, Assisi, Italy, October 22-24, 1993.

Key Words: Family structure, Moral disengagement, Real family, Ideal family.

R. G. Ardone, S. Vicini: Facoltà di Psicologia - Dipartimento dei Processi di Sviluppo e Socializzazione - Università di Roma.